DOP063. Biomarker panel for prediction of mucosal healing in patients with Crohn's disease under infliximab therapy
M. de Bruyn1,2, T. Bessissow3, T. Billiet2, I. Cleynen2, R. Kirkland4, X. Liu4, S. Hauenstein4, K. Drake4, S. Singh4, M. Ferrante2, P. Rutgeerts2, G. Van Assche2, I. Arijs2, G. Opdenakker1, S. Vermeire2, 1Rega Institute for Medical Research, Laboratory of Immunobiology, Microbiology and Immunology, Leuven, Belgium, 2KU Leuven, Department of Clinical and Experimental Medicine, Translational Research Center for GastroIntestinal Disorders (TARGID), Leuven, Belgium, 3McGill University, Division of Gastroenterology, Montreal Quebec, Canada, 4Prometheus Laboratories, Department of Research and Development, San Diego, United States
Infliximab has led to new therapeutic goals in Crohn's disease (CD) such as complete mucosal healing and improvement of quality of life. The current standard for assessing mucosal healing is endoscopy. However, frequent assessments are costly and uncomfortable to the patient. Non-invasive, accurate surrogate serum markers would therefore be welcomed to aid clinicians in predicting mucosal healing.
In a retrospective study, 119 CD patients who started infliximab and who underwent serial endoscopies (before and during infliximab) were included. Serum samples were available at the time of endoscopy and levels of markers were correlated with the degree of healing (not healed, marked improvement or complete healing). Thirty-five biomarkers were measured in 181 serum samples with the use of CEER, a proprietary highly sensitive protein micro-array, or homogenous mobility shift assays (Prometheus Laboratories Inc., San Diego, CA). These markers included growth and repair factors, pro- and anti-inflammatory markers and the IBD SGI serology panel. Infliximab and antibodies to infliximab were also measured. Clinical information regarding age, gender, age at diagnosis, location of disease, anal involvement and previous surgery was included. For statistical analysis SPSS and R were used.
From the 119 CD patients, 64 CD patients showed complete healing with no relapse, whereas 55 CD patients never showed mucosal healing. Univariate analyses indicated that age and 12 serum markers (HGF, BTC, TWEAK, CRP, ICAM, SAA, VCAM, IL-2, IL-8, IFN-γ, IL-6 and IL-10) were significantly associated with mucosal healing (p < 0.1). Selection of the best subset of predictors through multiple logistic regression analysis retained age [Odds ratio 0.97 (95% confidence interval 0.94–0.99), p = 0.010], HGF [0.86 (0.79–0.94), p = 0.001], BTC [1.24 (1.07–1.43), p = 0.003], TWEAK [1.04 (1.01–1.07), p = 0.014] and VCAM [0.93 (0.87–0.98), p = 0.012] based on Bayesian information criterion (BIC). The results remained significant after internal bootstrap validation. A cumulative prediction score was developed by combining 5 prediction factors (Age <29.5 years, HGF <11.42 CU/ml, BTC >11.44 CU/ml, TWEAK >20.62 CU/ml and VCAM <4200 µg/ml). Based on this cumulative prediction score, we could observe a significant and gradual increased prediction of mucosal healing (linear-by-linear p < 0.001).
We have identified a surrogate marker panel consisting of four serum markers and one clinical parameter that could facilitate prediction of mucosal healing in the future.